A Ten-Gene Transcriptomic Biomarker Panel for Glioma Classification and Prognosis Identified via Integrative Hypergraph and Rough Set Analysis

Simple Summary Glioblastoma is the most aggressive form of brain tumour, while lower-grade gliomas progress more slowly and respond differently to treatment. Distinguishing between these two categories early and reliably is essential for patient management, yet most molecular tests in current use require specialised platforms that are not universally available. In this study, we developed a computational method that combines network-based analysis of gene co-expression with a mathematical optimisation technique to identify a small panel of ten genes capable of separating these tumour types. The panel was trained on a Dutch microarray cohort and validated on two independent Chinese RNA-sequencing cohorts, retaining its accuracy across both technologies and patient populations. The same ten-gene panel also predicted patient survival independently of standard tumour grading, suggesting that it could complement existing clinical assessments and inform individualised treatment planning.